what we can learn and how we should do itwhat we can learn and how we should do it tom oosterloo...

15
What We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astronomy, Dwingeloo, NL Kapteyn Institute, Groningen, NL NGC 6946 H I (WSRT) and optical (DSS)

Upload: others

Post on 27-Aug-2020

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

What We Can Learn and How We Should Do It

Tom OosterlooNetherlands Institute for Radio Astronomy, Dwingeloo, NLKapteyn Institute, Groningen, NL

NGC 6946 H I (WSRT) and optical (DSS)

Page 2: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

What can we learn and how we should do it

H I studies are an important part of astronomy, but progress is slowing down

What are the opportunities offered by the new instruments?

What is the role of the EVLA

Page 3: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

Cannot understand the evolution of galaxies without knowing about their H I

Heavens et al.

Zwaan 2006

0 2 4

2

0

redshift

ΩH

I x 1

0-3

- The SFR density is a factor 10 higher at z = 1. Cosmic H I density does not track SFROn cosmic scales H I not 1-to-1 connected to evolution of star formation. Why???

Short gas consumption timescales. Accretion needed.How does this depend on environment, mass, type, redshift?

- In a cosmological context, we do not understand the connection between gas content and star formation and on what it depends, nor the evolution of this connection.

Doyle & Drinkwater 2006

Page 4: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

- We don’t understand the connection gas ⇄ formation, we also don’t know where

the gas is coming from!

- There is not enough H I at high z to form all stars in the current Universe. Closed box cannot work

- Only 10% of baryons are in galaxies, the rest is “out there”. What is the role of (cold) accretion?

Connected to this: how do galaxies get their gas?

Zwaan 2006The connection between star formation, H I and accretion, over time.is one of the main issues to address in the coming years through large, deep surveys of the H I in the local and “distant” Universe

Page 5: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

Gas is not accreted by eating dwarfs

- Galaxies have to accrete H I continuously

But:

- Most H I in galaxies is already in large galaxies

- Large galaxies cannot remain gas rich by accreting gas-rich dwarfs

CVn survey; Kovac et al 2007

Was accretion of dwarfs ever important?How does mass function evolve with redshift?

Page 6: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

H I halos are the interface between galaxies and IGM?

Counterrotating cloud

- Lower halo corresponds to galactic fountainbut outer halo has no kinematic relation to disk

- Can explain outer halo by accretion of gas from IGM- Only very few cases known because deep observations are needed- Deep H I observations of nearby spirals may help to understand the relation with IGM- Study the low column density H I in nearby Universe to see how this connects to IGM

extended non-rotating outer halo at level of 1019 cm-2, more than 20 kpc above disk

Oosterloo et al 200716x12 hr WSRT

NGC 891

Model of cold accretion(Maccio, Moore & Stadel)

Page 7: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

How we should do it?

Several new instruments available in a few years which will focus on H I

EVLAApertifASKAPMeerKatATAFast

Important to focus on what an instrument can do best, leave the rest to others

Page 8: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

Apertif (NL): 14 dishes of 25 m, focal plane array, FoV 9 degree2

MeerKat (ZA): 80 dishes of 12 m, single feed, FoV 1 degree2

ASKAP (AUS): 36 dishes of 12 m, focal plane array, FoV 30 degree2

All have bandwidths of ~300 MHz and 16000 (or more) channelscover H I over z=0-0.2 in one observation with good spectral resolution (few km/s)

Relevant for EVLA: these are survey instruments, a project of 1000 hrs is considered small

New facilities at L band in 2012

EVLA Apertif MeerKat ASKAP WSRT

A/T 2 1 1.3 0.5 1

FoV 1 30 4 120 1

Bandwidth 3 2 2 2 1

Survey Speed Shallow 4 30 7 30 1

Survey Speed Deep 12 60 14 60 1

SB Sensitivity + o o o o

Page 9: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

Latest result: first fringes on astronomical source between WSRT dish with FPA and “old” dish

Page 10: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

Apertif , ASKAP, MeerKat- Major new opportunity: can image the entire sky at high resolution, high sensitivity and out to

large distances

- Current state: we know about H I in ~104 galaxies, ~100 above z = 0, mostly single dish (3’-15’)

- In 2014: 106 galaxies, out to z = 0.6, most above z = 0.1, with 15 arcsec resolution.

- Can start addressing evolution of H I in galaxies and connection to star formation evolution

In parallel with- Can detect 105 M⊙☉ out to 5 Mpc. Over entire sky, so enough volume to study the smallest H I

galaxies, and smaller

- H I absorption survey out to z = 1. ~1000 detections

Page 11: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

What should the EVLA (not) do?

EVLA Apertif MeerKat ASKAP WSRT

A/T 2 1 1.3 0.5 1

FoV 1 30 4 120 1

Bandwidth 3 2 2 2 1

Survey Speed Shallow 4 30 7 30 1

Survey Speed Deep 12 60 14 60 1

SB Sensitivity + o o o oNiches for EVLA:

- Very deep integrations of small fields. Go for highest z- Deep integrations of nearby spirals. Interface galaxies - IGM- Deep searches for low column density H I in nearby Universe with compact array configurations

Only worth doing if willing to spend large amounts of observing time

note: 100 h with EVLA = 1 min with SKA

Page 12: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

Optimum resolution for deep H I work

Column density averaged over disk is the same for all spirals so you can compute at which resolution you start resolving out galaxies at high z

Broeils & Rhee 1997EVLA 100 hrs 100 days

z = 0 4”-7” 2’’-3’’

z = 0.2 5”-9” 2”-4”

z = 0.4 7”-13” 3”-5”

z = 0.6 9”-17” 4”-8”

See ASKAP memos of Staveley-Smith for more details

No piggy-back with continuum surveysH I integration times too long

size of smallest galaxy

Page 13: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

“The best place to find a galaxy is right next to another one”

This is not good!!!!

Galaxies are clustered so confusion is potentially a problem

Simulations done (for ASKAP) by Lister Staveley-Smith

For deep C-array surveys:

Need to have independent redshift information, otherwiseconfusion can be significant

Photometric redshifts are sufficient

Confusion by galaxies > 0.1 M ★

By galaxies > 1 M ★ is factor 15 down

Staveley-Smith 2008

Page 14: What We Can Learn and How We Should Do ItWhat We Can Learn and How We Should Do It Tom Oosterloo Netherlands Institute for Radio Astr onomy , Dwingeloo, NL Kapteyn Institute, Gr oningen,

SummaryImportant issues to address in extra-galactic H I work

- One cannot understand the evolution of galaxies without knowing about the evolution of H I in galaxies

- Investigate the H I properties as function of type, environment, mass, …In particular, the evolution of the above with redshift and the relation with evolution of star formation

- How do galaxies get their H I, interface between galaxies and IGM

- Low column density H I, Cosmic Web

- The smallest galaxies, and smaller

- Few niches for EVLA

• Deep integrations of small fields. Go for highest z• Deep integrations of nearby spirals. Interface galaxies - IGM. Combine with high z?• Deep searches for low column density H I in nearby Universe with compact array configurations• Also EVLA projects will need very long integration times